DE2854843A1 - DATA TRANSFER SYSTEM - Google Patents

Description

The invention relates to data transmission systems and in particular to such DaLon transmission systems, who make more effective use of available telephone transmission lines than systems previously known in the art.

Because of the extremely high cost of certain telephone transmission lines, e.g. satellite transmission lines and submarine transmission lines, one has tried so far by various means to improve the effectiveness and performance of existing transmission lines to make it as big as possible. One such system is as a time-dependent speech interpolation system or TASI system known. In a typical TASI system, calls or connections from N Transmit input lines or channels to a remote location via, for example, N / 2 transmission channels. At this location, the N / 2 transmission lines are in turn connected to N output lines or channels. TASI systems are based on the assumption and assumption that are verified as statistical fact means that not all channels will be in use at any given time. In fact it is considered general Rule that the lines or channels are actively in operation in less than half the time. Accordingly TASI systems can be defined as switching or selector systems, the transmitter and receiver only then with one another connect when the transmitter is active, provided that there is a transmission facility or transmission line is available at this time.

A problem that occurs with the known TASI systems is the difficulty of so-called freezing out. Freezing out is a failure or loss problem

909826/0858

language or other signals that can occur whenever the number of people working in a Direction speaks or begins to speak, exceeds the number of available transmission lines. The greater the number of transmission lines is, the less likely is the chance that a particular caller will not have a transmission line assigned will when he needs one. An attempt to interpolate two independent conversations on a single channel, would thus lead to the fact that a large percentage of the conversation is lost because the statistical probability that both participants speak at the same time would be very high. With a big one However, the group or number of transmission devices which form an even larger group of channels is the fluctuation in demand for transmission equipment or lines is much smaller than a proportion of the total capacity, and thus freeze-out is reduced. Well-known TASI systems can, for example Use 37 or more transmission facilities to reduce the statistical likelihood of freezing out.

It is desirable to apply the principles of time-dependent speech interpolation methods to systems in which which a smaller number of transmission facilities is available. For example, it may be desirable be the TASI principle on private transmission line systems to transmit, which have only four transmission facilities.

In US-PS 4,012,595 an attempt is already given to reduce the number of transmission devices,

909826/0858

which is required in a digital transmission system are. The system described in US Pat. No. 4,012,595 includes an encoder to encode the original speech signals to convert it into digital form, as well as a digital voice memory to store these voice signals temporarily in the order in which they appear. The language is then retrieved from the voice memory and transmitted in digital form. The storage of the language is very short-term and depends on the Switching or switching process that is used to convert the speech blocks into transmission times in the System.

A problem encountered with that described in U.S. Patent 4,012,595 System occurs is that the transmission of voice signals in digital form is a relatively large Requires bandwidth and thus performance of the system decreased. While the system described there with the use of only 15 transmission channels To be effective, it is desirable to have a system for use in home telephone line systems that for example only four transmission facilities owns.

Efforts have also been made to reduce the amount of memory required in TASI systems. For example, US Pat. No. 3,684,680 deals with the storage capacity in a TASI system to keep it to a minimum. Accordingly, only those speech signals are stored there that Transmission channels have been assigned. In the system described there, the storage capacity is kept to a minimum, however, there is

- 10 -

909826/0858

- 10 in no way solved the problem of freezing out.

Accordingly, the invention is based on the object of creating a telephone data transmission system, that is able to handle a relatively large number of callers over a relatively small number of transmissions s lines to connect with a large number of listeners.

The telephone data transmission system according to the invention is characterized by the fact that the freezing out or the loss of messages despite a relatively low Number of transmission devices can be kept at a minimum value.

A TASI system is used in the system according to the invention operated with a small number of transmission facilities without freezing or message loss by the sampled input signals are stored for variable times until a transmission line or device available.

In particular, the data transmission system according to the invention has a sampling device that provides a digital representation of an incoming signal. If no If transmission channels are available, this sampled signal is stored in a voice memory for as long stored until a free and available transmission channel can be found. If such a Transmission channel is available, the stored and sampled signal is transmitted by means of a control device retrieved from memory and, in one embodiment, reconstructed again to an analog signal.

- 11 -

909826/0858

This reconstructed signal is then transmitted to a remote location. Such a system is so far advantageous than the transmission of the reconstructed speech signal in analog form, which leads to a Profitability and savings in bandwidth leads.

Another advantageous property of the invention System is that a device is provided to store the above-mentioned voice mailbox to be circumvented if a transmission facility is available and if none previously occurred Voice signals have been stored. The provision of such a bypass device enables Transmission of signals in real time, if not a situation of freezing out in which all Lines are busy occurs.

The invention is explained below with the aid of the description of exemplary embodiments and with reference to FIG attached drawing explained in more detail. The drawing shows in

Fig. 1 is a block diagram of a terminal station of the entire system according to the invention;

Fig. 2 is a block diagram to explain the

Transmission part of the connection station of the telephone data transmission system according to FIG. 1; 30th

FIG. 3 is a block diagram of an embodiment of FIG fixed and variable buffers of Figure 1;

- 12 -

909826/0858

4A and 4B are graphs for explaining the timing sequence used in an embodiment of the system of the present invention;
5

5A and 5B are flow charts for explaining the operation the control device according to FIG. 2; and in

6 shows a block diagram to explain the receiver part of the telephone data transmission system according to the invention.

Figure 1 shows a block diagram of one end of the overall system which is the subject of the invention. A subscriber device 1, which is a subscriber exchange or a transmitter device can be used by a private line data transmission system operated, which has four-wire transmission devices 11. The participant's device is via input channels 10 connected to the system of which the system according to the invention forms a part. The N input channels are normally served by approximately N / 2 transmission facilities, usually up to 31 Input channels supplied by up to 16 transmission devices.

Interface lines 2 of the channels are with analog-to-digital converters and a time divider circuit to periodically poll each of the input channels or to ask to look after them one after the other. Fixed and variable transmission buffers 3 are also provided. In the new system, the variable buffers provide temporary storage of the scanned data

- 13 -

909826/0858

Signals when no transmission device is available stands. A symbol generator 4 generates a tone symbol which is transmitted in front of each via a transmission device Speech block is used to indicate the channel that delivered the speech block when the Speech block uses a transmission facility that is not already assigned to the channel. The fixed transmission buffer 3 provides a time interval in which the sound symbol is to be transmitted without the input signal to cut off.

The speech blocks are connected to the transmission devices 11 via the interface lines 5. The interface lines 5 have a time divider, a digital circuit arrangement and in one Embodiment digital-to-analog converter. The output signal this digital-to-analog converter is an analog signal, and this analog signal is transmitted via the transmission equipment transmitted to a remote OrL, sometimes referred to as a "remote end of the line" will.

When speech blocks or information are received or received from the remote location, the receiver buffers supply 6 with a fixed length, a time interval in which the symbol detector 7 decodes the sound symbol to to determine which channel 10 the information is to be assigned to. The assignment of channels to transmission facilities and the time during which an information or speech block is in the variable speech buffer can be stored, is subject to the control of the controller 8, which is usually a Microprocessor acts.

- 14 909826/0858

Reference should now be made to FIG. 2, in which the transmission part of the data transmission system in the individual is shown.

The output or original _{signals V Q} are applied to the plurality of input channels 1, 2 of the type shown in Fig. 6 can be received. The transmission devices are denoted by 11. The number N / 2 of transmission channels can reach a value of 16, for example, but one of the advantages of the system according to the invention is that the number of transmission channels N / 2 can be much smaller than this number, for example the value 4, without that The problem of excessive freezing outlined above occurs.

In order to transmit the signals arriving on N input channels via approximately N / 2 transmission devices, each of the input channels 1, 2 ... N is connected to the interface or the interfaces 2 of the channels, which have a large number of scanning devices, for example analog-digital Converters 12, which are released sequentially by a multiplexer 13. The analog-to-digital converters 12 are preferably coding-decoding devices (CODEC devices) which supply pulse-code-modulated blocks or PCM blocks. Each of the sampling devices in the form of the analog-to-digital converter 12 generates a digitally coded signal V ₁ , which is a digital representation of the original analog signal V _Q. The input lines

- 15 -

909826/0858

1, 2 ... N are also as shown each to a speech detector 14. The speech detectors 1 4 speak to the presence of an original signal Vq on an input channel and are able to switch between speech and distinguish most random fluctuations or noise. The output of each speech detector 14 is connected to a control module or controller 8, which is preferably a microprocessor acts. The controller 8 arranges one of the transmission devices in a manner similar to that in other TASI systems 11 to a special input channel if a transmission device 11 is available. Accordingly a memory 22A for transmission possibilities is called up in order to determine which transmission device, if at all, is available. The controller 8 can be a commercially available microprocessor, for example an Intel 8085Ά.

Each of the sampled signals V- is applied to a delay device 17 designed for a fixed time, for example a fixed transmission buffer, whose function does not need to be explained in more detail here. When the sampled signals V- from the delay device 17 exit, each of these sampled signals or each block is sent to a switching device 18 created, the position of which is aligned by the controller 8. The controller 8 also controls the state of an additional switching device 19. The pair of switching devices 18 and 19 are each over a line or a detour line 20 connected to one another. At the moment when the memory 22A for transmission possibilities a condition indicates that a

- 16 -

909826/0858

Transmission device 11 is available when it is called up by the controller 8, the sampled signal V ₁ is fed through the action of the switching devices 18 and 19 via the detour line 20 of the relevant transmission device. The switching devices 18 and 19 and the detour line 20 thus form a device for bypassing the variable speech buffer 22.

The switching devices 18 are also connected via lines 24 to the voice buffer 22 for variable length. At the moment when the memory 22A for transmission options indicates that none of the transmission devices 11 is available when they are called up by the controller 8, the controller 8 actuates the switching device 18 and forwards the coded or sampled voice signal V _{1 to} the voice buffer 22, where the encoded signal is temporarily stored until a transmission channel is available. The coded speech signals V ₁ , including short periods of silence, are stored in the speech buffer 22 in the order in which they occur. Furthermore, all coded speech signals V ₁ that originate from a particular input channel are stored within the speech buffer 22 at the same location. When a transmission device 11 actually becomes available, these stored signals are called up from the memory by the controller 8 by activating the switching device 19.

Regardless of whether the sampled signal V _{1 has} been stored in the voice memory or has been routed around the memory via the detour line 20, the signals V ₁ after passing through the switching input

- 17 -

909826/0858

direction 19 is applied to the interface circuit 5 via a circuit element 21 with a holding function (latch). The interface circuit 5 is provided with a switching circuit 26. This circuit 26 connects the coded voice signals V- to the free transmission device 11. According to a very essential aspect of the present system, however, the signals V _{1 are} first reconstructed by a plurality of digital-to-analog converters 28, one of which is provided for each transmission device 11. The output signals of these digital-to-analog converters 28 are reconstructed signals VA, which are an image of the original signal V ₀ . The reconstructed signals V1 are then transmitted in analog form to a receiver circuit, and this analog transmission is advantageous in that analog signals enable the transmission bandwidth to be used economically.

The controller 8 is provided with a timing device for registering the time interval between the arrival each successive signal V from a particular line of the input lines to speech buffer 22 Mistake. The control then calls the stored signals in chronological order of their arrival time at the voice buffer 22. Signals saved later are called up one interval later than earlier stored signals, the interval being equal to the time interval between their arrival times. on in this way the signals are removed from the speech buffer 22 without significant distortion or change in order and duration of the speech blocks retrieved.

According to a further essential aspect of the novel system, the above-mentioned bridging or switching

- 18 -

909826/0358

Path arrangement, which consists of the switching devices 18 and 19 and the detour line 20, is advantageous in that not all of the incoming signals need to be stored in the voice buffer 22. _{If the coded voice signals V 1} arriving on a special input line 1, 2. However, before this signal can bypass the voice buffer 22, the controller 8 must first scan and determine whether any previously received voice signals V- are present in the voice buffer 22. If this step were not carried out, it would be possible that coded speech signals V ₁ would be transmitted in an incorrect time sequence. Thus, all incoming signals are fed from a special input line to the voice buffer 22 unless a transmission device 11 is available and no previously stored coded voice signals V ₁ originating from this special line are present in the voice buffer 22.

The control module or controller 8 maintains the sequence all speech blocks on the respective input channel, so that no overlapping speech blocks can be generated by an input channel via two transmission devices at the output. An example of a Such an overlap can, for example, be a first speech block from the input line 1 in the speech buffer 22 are stored and wait for a free transmission facility. If a second language block appears occurs on this input channel, two transmitted

- 19 -

909826/0858

facilities are free and available, so that it is conceivable that the second speech block can be transmitted at the same time as the first speech block " te. This problem of overlapping speech blocks is avoided in the present system because the control 8 as long as no transmission device is available for a certain input channel until all «- Liche speech blocks from this channel, which have been stored in the speech buffer 22, completely via a transmission device have been sent out. In this way the second block of speech always becomes sequential after the previous block of speech on the same input line sent out.

Referring to Figure 3, a specific embodiment of the fixed delay device 17 and the voice buffer 22 variable length are described in detail.

In the embodiment according to FIG. 3, the incoming sampled voice signals V ₁ from a special input channel 1, 2,... N are applied to a high-speed buffer 15 in serial format. These sampled signals V- are then retrieved from buffer 15 and applied to direct access memory or RAM memory 200 in a parallel format. Instead of using a fixed and a variable buffer separately, the best way of realizing the novel system is to / use a single RAM memory / which has both a fixed buffer in a fixed part 170 and a variable buffer in a variable part 220 in this memory. However, it is also within the scope of the invention to

- 20 -

309826/0868

separated to use fixed and variable delays or transit times, as in the embodiment according to Fig. 2 has been specified in detail. It is also within the scope of the invention, the input signals sample and store analog blocks in an analog memory such as a delay line or a load transfer device.

At the moment when one of the transmission devices 11 is available and no previously occurring, sampled signals V ₁ from a special input channel are stored in the variable part 220 of the RAM memory 200, the one originating from this special input channel and in the fixed part 170 of the RAM memory stored, scanned signal V _{1 is} transmitted with 8 parallel bits to a circuit element 21 with a hold function. In addition, the most recently sampled signal V _{1 is} transmitted from the buffer 15 to the part 170 of the RAM memory 200 with a fixed delay.

If, however, one of the transmission devices 11 is free and previously sampled signals V that have occurred are stored in the variable part 220 of the RAM memory 200, the oldest stored sampled signal V _{1 is} transmitted from the variable part 220 to the circuit element 21 with a hold function. In addition, the last sampled signal V ₁ , which is to enter the fixed part 170 of the RAM memory 200 from this channel, is transmitted to the variable part 220. To make a transfer from the fixed part 170 to the variable part 220 of the RAM memory 200, the content of the fixed part 170 of the RAM memory 200 is in a

- 21 -

909826/0858

fast buffer 230 is emptied, where or readdressed and written into the variable part 220. In addition, the sampled signal V ₁ generated very last is transmitted from the buffer 15 to the fixed part 170 of the RAM memory 200.

At the moment when no transmission device is free, the sampled signal V ₁ stored in the fixed part 170 of the RAM memory 200 is transmitted to the variable part 220 via the high-speed memory or the high-speed buffer 230. In the unlikely moment when no transmission device 11 becomes free before the variable part 220 has been completely filled, the memory will be overwritten, with the result that speech blocks are lost.

The time sequence of the operations and steps specified in connection with FIG. 3 will be explained in more detail below with reference to FIGS. 4A and 4B. To explain the two FIGS. 4A and 4B, it is sufficient to assume that N = 31, ie that 31 separate input channels are present. For example, if each of the input channels is fetched from multiplexer 13 once every 125 microseconds, data will arrive at RAM memory 200 from the various input channels about once every 4 microseconds (125 microseconds divided by 31 input channels). This time frame or the time interval of 4 microseconds during which a special input line is called is further subdivided into eight sub-intervals, namely the sub-intervals a _Q , a ₁ , b _Q , b ₁ , c _Q , c ^, shown in FIG. 4B. d _Q and ä ^, each of these sub-intervals being approximately 0.5 microseconds in length.

- 22 -

909826/0858

During the sub-interval a _Q , if a transmission device 11 is available and if there are no sampled signals V ₁ from the special input channel just called up, for example input channel 3 in the illustration according to FIG. 4A, in the variable part 220 of the RAM memory 200 are stored, the sampled signals V ₁ of this special channel are transmitted from the fixed part 170 of the RAM memory 200 to the circuit element 21 with a latch function. Furthermore, during this sub-interval, the most recently detected scanned signal V _{1 is} transmitted from the input channel just called up from the buffer 15 to the fixed part 170 of the RAM memory 200.

During the sub-interval a-, when scanned signals V ₁ from the special channel just called up are stored in the variable part 220 of the RAM memory 200 and when a transmission device 11 is now free, the oldest scanned signal V ₁ from the variable part 220 of the RAM memory 200 transferred to the circuit element 21 with a hold function. A sampled signal V _1, which is stored in the fixed portion 170 of the RAM memory 200 is transmitted via the fast buffer 230 to the variable portion 220 is then during this interval part a ... In addition, the sampled voice signal V ₁ that occurred most recently is transmitted from the buffer 15 to the fixed part 170 of the RAM memory 200 during this time interval.

Finally, during the sub-interval b,. _{transmit any signal V 1} present on the circuit element 21 with a hold function to an output transmission device 11.

- 23 -

909826/0858

The remaining sub-intervals C _n , C ₁ , d- and d ₁ are used by the control module or the controller 8 to carry out the decision processes mentioned above. The various decisions made by controller 8 can best be explained in conjunction with the flow chart of Figures 5A and 5B. As shown in Fig. 5A, after a sampled voice signal V _{1 has} been collected from a particular input _{channel, the controller 8 reads the most recent sampled signal V 1} from that channel, which is stored in the fixed part 170 of the RAM memory and writes the current signal to the same address. As illustrated by decision block 310, the controller 8 then determines whether an output transmitter 11 has been assigned to that input channel. When a transmission device has been assigned, that signal V- which has been stored for the longest time is transmitted to the circuit element 21 with a hold function. If, however, no transmission device 11 has been assigned, the controller 8 determines whether the user of the particular input channel in question is speaking. This is indicated by decision block 312. If there is no speech on the input line, control determines whether the channel is assigned to a variable buffer in accordance with decision block 314A. If this is not the case, the control switches on and calls up the next channel. If so, control passes to decision block 318, discussed further below. If the input channel is spoken, the controller 8 determines whether space has been allocated to this input channel in the variable part 220 of the RAM memory 200. This is due to the decision

- 24 -

909826/0858

indicated block 314B. If no room is assigned if i:; t, this will be taken care of. If, however, space has been allocated beforehand, then the control system restores 8 determine whether this space is expanding. 5

In other words, the controller 8 determines whether the channel X assigned to the buffer has received a transmission option, and if this is the case, so it does not expand any further. This is illustrated by decision block 316. If the participant pauses in a conversation, which is sensed by the controller in accordance with decision block 312, so provides the controller 8 determines whether the subscriber's silent period is longer than the time required by the subscriber Read out participants stored speech blocks. This is illustrated by decision block 318. The implementation The decision represented by decision block 318 thus eliminates the above mentioned overlap problem of speech blocks.

At the moment when the controller 8 detects that the allocated buffer expands, i.e. no transmission device is allocated to the buffer, so the controller determines whether there is more space available in the variable part of the RAM memory 200. this is indicated by decision block 320. If there is space available in the variable part 220, it will stored sampled speech signal V. from the fixed part 170 to the variable part that occurred very recently of the RAM memory 200 is transferred.

The information stored in the fixed or variable parts of the RAM memory 200 is transferred to the latch or circuit

- 25 -

909826/0858

element 21 supplied with a hold function, from where it is in turn transmitted to the receiver part of the novel system.

In Figure 6 the receiver portion of the novel system is general designated by 30. The receiver part 30 represents a device to receive the reconstructed signals V ', which are transmitted via the transmission device 11 have been.

The receiver part 30 has a multiplicity of scanning devices 32, for example in the form of the analog-to-digital converter shown. The scanning device 32 preferably provides a re-encoded speech signal, V ₁ ^1, which is a digital representation of the reconstructed speech signal Vl. The re-encoded signal V 'is then applied to a further delay device 34 with a fixed time, for example a fixed receiver buffer memory. The delayed, re-encoded signal V ′ is then applied by a second controller 36 via a multiplicity of reconstruction devices 40 to the selected output channels or output lines of a multiplicity of output lines 38. As can be seen from the drawing, the reconstruction devices consist of digital-to-analog converters. The output signal of the reconstruction devices 40 represents an image of the original speech signal V _n which has been applied to the input channels 1, 2 ... N, and these signals are applied to the output channels 38 and supplied to the respective telephone subscribers. The controller 36 responds to the control information in order to assign the respective transmission channel to the correct output channel 38.

909826/0858

Claims (1)

PATENT LAWYERS DR. KADOR & DR. KLUJNKER K 12 STORAGE TECHNOLOGY CORPORATION 2270 South 88th Street Louisville, Colorado 80027 USA Data transmission system Claims Π 3 data transmission system, with a device for Transferring input signals from a plurality of input channels to a circuit through a circuit lower number of transmission facilities that are directly connected to a remote location, marked - By speech detectors (14) which are each connected to an input channel (10) to detect the presence of an input signal (V _Q ) on this channel (10) to be scanned, - By means (12, 13) for interrogating the scanned Signals (V.,) on the respective input channels (10), - Through a voice buffer (22) of variable length to the temporary storage of the scanned signals (V ₁ ) / if no transmission device (11) is free until such a transmission device (11) is available, - by one connected to the speech detectors (14) Control (8), which the sampled speech signals 909826/0858 " ² - ORIGINAL INSPECTED (V-) from the respective input channels (10) either via the circuit (18-20) available transmission devices (11) or which assigns these signals (V ₁ ) to the voice buffer (22) if there is no transmission device (11) for Is available, and which then _{sequentially retrieves the sampled signals (V 1} ) from the voice buffer (22) when a transmission device (11) becomes free, and - by one connected between the circuit (18-20) and the transmission devices (11) Device (26, 28) which _{forms reconstructed speech signals (Vl) from the sampled signals (V 1} ) which are applied to one of the transmission devices (11), 2, data transmission system according to claim 1, characterized by a device (18-20) which is connected to the controller (8, 22A) and which guides the sampled signals (V ₁ ) coming from a special input channel (10) around the voice buffer (22), when both a transmission device (11) is available and no sampled signals (V ₁ ) from this channel (10) are stored in the memory (22). 3, data transmission system according to claim 2, characterized in that the device (18-20) has a pair of interconnected switching devices (18, 19) for bypassing the voice buffer (22). 4. Data transmission system according to claim 3, characterized in that the pair of interconnected switching devices (18, 19) from the controller (8) is controlled. 909826/0858 5. Data transmission system according to claim 1, characterized in that the voice buffer (22) variable length is active long enough after the coded voice signal (V ₁ ) has been transmitted to the transmission device (11) in order to overlap speech blocks impede. 6. Data transmission system according to claim 1, characterized in that the controller (8) is provided with a time measuring device for registering the time intervals between the arrival of the respective successive scanned signals (V ₁ ) from a special line of the input channels (10) at the voice buffer (22) is. 7. Data transmission system according to claim 6, characterized in that the controller (8) has a device which _{retrieves the stored sampled voice signals (V 1} ) in the chronological order of their arrival time at the voice buffer (22), with signals stored later being retrieved later than previously saved signals. 8. Data transmission system according to claim 1, characterized in that g e <■ * indicates that the device (12, 13) for calling up and sampling the signals analogue-to-digital converter (12). 9. Data transmission system according to claim 1, characterized in that that the device (26, 28) . for generating reconstructed speech signals [Y ') digital-to-analog converter (28). 909826/0858 10. Data transmission system according to claim 1, characterized by a device (30) for receiving the first reconstructed voice signals (V ¹ ) from the respective transmission devices (11), this receiver part (30) having a device (32) for re-digitally coding the first reconstructed speech signals (Vq), a second controller (36) for selectively applying the first reconstructed speech signals (VI) to a selected output line of a plurality of output lines (38) and means (40) for reconstructing the original or input signals (V _Q ) the re-encoded signals (V ') on the output channels (38) is provided. 11. Data transmission system according to claim 10, characterized characterized in that the receiver part (30) is fixed with a delay device or a buffer Length (34) is provided to store the re-encoded speech signals (VJ) until the assignment of the Management has taken place. 12. Data transmission system, with a device for transmitting signals from a plurality of input channels via a switching device to a lower one Number of transmission devices, which with a remote location are marked - by speech detectors (14) which are each connected to an input channel (10) in order to sense the presence of speech signals (V ₀ ) on the channels (10), - by a device (12, 13) for retrieving the sampled voice signals (V-) from the respective input channels (10), 909826/0858 - by a voice buffer (22) for temporarily storing the coded voice signals (V ₁ ) / if no transmission device (11) is available until a transmission device (11) is free, - by one to the voice detectors (14) and the voice buffer ( 22) connected controller (8) which _{either assigns the scanned signals (V 1} ) to the transmission devices (11) available via the switching device (18-20) or assigns the voice signals (V ₁ ) to the voice buffer (22), if no transmission device (11) is free and which then _{sequentially calls up the stored, scanned signals (V 1} ) when a transmission device (11) becomes free, and - by a device (8, 22A) which _{routes the sampled signals (V 1} ) from a special input channel (10) around the voice buffer (22) when both a transmission device (11) is available and no sampled signals ( V ₁ ) from the respective channel (10) are stored in the speech buffer (22). 13, data transmission system according to claim 12, characterized in that the control is provided with a time measuring device for registering the time intervals between the arrival of the respective successive signals (V ₁ ) from a special line of the input channels (10) at the voice buffer (22). 14. A data transmission system according to claim 12, characterized in that the controller (8) has a device which receives the stored sampled voice signals (V ₁ ) from a special input channel (10) in chronological order of their arrival time 909826/0858 at the voice buffer (22), with signals stored later being retrieved later than previously stored Signals at time intervals that are only due to the actual availability of transmission facilities is determined. 15. Data transmission system according to claim 14, characterized in that the coded voice signals (V ₁ ) from a special input channel (10) are stored until a transmission device (11) is available and until all previously occurred, stored sampled signals (V ₁ ) are retrieved from this channel (10). 16. Data transmission system according to claim 12, characterized characterized in that the means (18-20) for bypassing the voice buffer (22) is a pair of each other has connected time divider and space shift switching devices, 17. Data transmission system according to claim 16, characterized characterized in that the switching device can be actuated by the controller '(8). 909826/0858